Yes, I'm sure we are all aware that numbering things enables you to place them 
in order.  No doubt you see the relevance of block numbers within a dataset 
written as part of a block on DASD to some order of network transmission 
between DASD devices which quite escapes me.


-----Ursprüngliche Nachricht-----
Von: IBM Mainframe Assembler List [mailto:[email protected]] Im 
Auftrag von John Gilmore
Gesendet: Donnerstag, 13. September 2012 22:16
An: [email protected]
Betreff: Re: Data spaces vs hiperspaces

Consider a dataset comprised of n blocks, numbered 0, 1, 2, . . . , i,
. . . , n - 1.

Let us now send that dataset from some source s to some sink S over a
'network' that contains multiple paths from s to S.  Transmitted in
the canonical sequence 0, 1, 2, . . . , n  - 1, they may arrive in a
different sequence, in principle in any permutation of these n block
numbers.  The presence of a block sequence number (relative block
number) in each block [or block extension] then permits the
transmission sequence at s to be recovered from the arrival sequence
at S (if there are sufficient buffers available at S).

Note also that that what are buffered by access methods are blocks not
[logical] records and that with SAM-E the historically clearcut
distinction between BSAM and QSAM, if not abolished, was certainly
muddied.

John Gilmore, Ashland, MA 01721 - USA

Reply via email to